Talk:OLPC Human Interface Guidelines/The Sugar Interface/Colors
I don't really understand this use of color. It seems contradictory to say that you cannot use color like other UIs because the screen may be used in monochrome mode, but you are instead using color to identify a particular child. If color is significant, then how will it be useful in monochrome mode? In other words, how will a child identify another child's icons, text, etc. if the display is in monochrome mode?
Wouldn't it be better to use color for decoration and not for significance? That way, whether the display is in backlit color or reflective monochrome, there is no information loss.
- This is a delicate issue, and we've thought hard about how to balance it. While it's true that we "lose information" in monochrome mode, our use of the Munsell value scale will ensure that, even without chroma information, the various identities chosen by the kids remain distinguishable to the extent possible with fewer dimensions. Additionally, in some particular cases, names/photos may be used instead; for instance, in a chat activity. - Eben 15:08, 16 January 2007 (EST)
- Forget that Munsell stuff. It's too limiting, doesn't work very well, and forces many other choices. Go with patterns, especially ones that look like real-world texture. AlbertCahalan 00:38, 2 March 2007 (EST)
I should also mention that we've found that children (esp. preteens and teens) are constantly experimenting with their identities and changing their surroundings whenever possible. This means to me that kids will want to change their colors regularly, again reducing the value of color as an identifier. Since the XO has a built-in camera, why not use a combination of the child's name and photo to identify their work. Admittedly a tiny thumbnail can be hard to recognize, but I think that AIM, XMPP, Google Talk, etc. have shown the value of even a tiny image (which can also be zoomed when the mouse cursor hovers over it).
- This is also a very true statement. Indeed, we will be taking advantage of the camera and capturing a photo of each child with an XO laptop. Though this photo won't be shown everywhere on the interface that the child's colors are, it will always be visible in the rollover for their XO avatar (which is always present on the mesh, and also appears within each activity that they participate in). This gives us a way to map the colors for an individual to their name and face. Though this doesn't eliminate the problem of constantly changing identities (via colors), it does mean there will be a way to keep track of it. The main reason we chose to take this approach is because, on an already small screen, there is little room for countless little photos in every location we might want to identify an individual. - Eben 15:08, 16 January 2007 (EST)
I also find the current monochrome UI is not very appealing. If you want children to embrace and become attached to their laptops, it seems important to make the UI as visually interesting as possible. Again, kids love to customize their surroundings, so I would think that custom background images, colors, etc. would have strong appeal.
- Again, I am in agreement. However, there were calculated reasons for this as well. I think the biggest reason, aesthetically speaking, is that we wanted to highlight the nature of the truly unique mesh capabilities. The idea is that, though the laptop is bland by itself, the community fills it with color. This can be seen in current screenshots of the mesh view. Additionally, activity developers are strongly encouraged to incorporate the childrens' colors within the activities where applicable. For instance, text written by a child could be in their colors in a writing activity; or a photo taken by them; or the particular track in a music composing activity; or their gamepiece; etc. In this manner, the interface becomes filled with colors that (even if somewhat abstractly) represent the children themselves. Furthermore, the gray interface should make the content the kids create - one of the main goals of the project is creation - pop visually. The interface takes a back seat to the drawings, games, webpages, etc. that the children produce.
- Of course, as noted, I agree with many of your points. The fact is, we're taking some chances with this approach, and we're not sure just how successful it will be. We may have to introduce some colors, or colored backgrounds, or totally change some of these ideals, but we're going to push this idea to the limit first. - Eben 15:08, 16 January 2007 (EST)
Some of the Sugar Designs images caused be to think of this paragraph from Knut Inge Fosterwold's and Dennis R. Ankrum's chapter 4, Visual Ergonomics for Children, in Ergonomics for Children, Designing products and places for toddlers to teens, Edited by Rani Lueder and Valerie Berg Rice [ Taylor & Francis 2008 ISBN 978-0-415-30474-0],
Too many colors, however, can produce a "Christmas tree" effect, making it difficult to determine which information matters. This is especially important for children, as they lack the training and ability to handle complex information. Children pay more attention to the attractiveness of a multicolored message than to the content.
I have found the color saturation on the XO to be lower than the LCD I am using to review the images, so the distraction threshold may still be largely avoided. --FGrose 18:49, 4 March 2008 (EST)
There are some web resources, such as, Vischeck that might be useful for accommodating colorblind users. Web safe color recommendations for colorblind users from British Telecom Labs  and VisiBone.com . A sample of the differences  from BTLabs. See also Accessibility. --FGrose 15:57, 5 March 2008 (EST)
We might also want to consider ways to avoid the confusion or annoyance of false stereopsis, such as when blue text appears on a red background. Some of the XO color combinations tend toward that for me. --FGrose 18:25, 4 March 2008 (EST)